1 Illuminating the dark ages: the most distant quasars in (public) surveys Bram Venemans (MPIA)
2
Collaborators KiDS QSO team (Kapteyn) - Gijs Verdoes-Kleijn - Johnson Mwebaze - Edwin Valentijn + KiDS team (PI: Kuijken) VIKING quasar SWG - Richard McMahon (IoA) - Joseph Findlay (Wyoming) - Will Sutherland (QMUL) UKIDSS quasar group - Steve Warren (Imperial) - Daniel Mortlock (Imperial)
Team Quasars @ MPIA - Eduardo Bañados - Roberto Decarli - Ema Farina - Chiara Mazzucchelli - Fabian Walter - Laura Zschaechner Pan-STARRS1 KP10 - Ken Chambers (Hawaii) - Gisella De Rosa (Ohio) - Xiaohui Fan (Arizona) - Hans-Walter Rix (MPIA) - Rob Simcoe (MIT)
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
3
Distant quasars: probes of the early universe
• Quasars at redshifts z ≳ 6 can be used to:
- determine the state of the intergalactic medium - measure space density of massive black holes
- study the formation of massive host galaxies - locate galaxy overdensities in the early universe
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
4
The search for distant quasars
• Problem: high redshift quasars are very rare
→ not found in deep (HST) blank fields
→ need multi-colour surveys over large area
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
5
The search for distant quasars
• Problem: high redshift quasars are very rare
→ not found in deep HST blank fields
→ need multi-colour surveys over large area • SDSS very successful, discovered many luminous quasars up to z = 6.4
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
6
Fan, Carilli & Keating 2006
19 quasars at 5.7 < z < 6.4 from the SDSS survey
z = 5.7
z=6.4
9800 Å06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
7
The search for distant quasars
• Problem: high redshift quasars are very rare
→ not found in deep HST blank fields
→ need multi-colour surveys over large area • SDSS very successful, discovered many luminous quasars up to z = 6.4 → to find quasars at higher redshifts, wide field NIR surveys are needed
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
8
Quasar searches in near IR surveys
• UKIDSS LAS
• VISTA/VIKING
• Pan-STARRS
• Dark Energy Survey, HSC, Euclid, LSST, …
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
9
Quasar searches in near IR surveys
• UKIDSS LAS: 4000 deg2, JAB~20.5
• VISTA/VIKING
• Pan-STARRS
• Dark Energy Survey, HSC, Euclid, LSST, …
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
10
Quasar searches in near IR surveys
• UKIDSS LAS: 4000 deg2, JAB~20.5
• VISTA/VIKING: 1500 deg2, JAB~21.5-22.0
• Pan-STARRS
• Dark Energy Survey, HSC, Euclid, LSST, …
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
11
Quasar searches in near IR surveys
• UKIDSS LAS: 4000 deg2, JAB~20.5
• VISTA/VIKING: 1500 deg2, JAB~21.5-22.0
• Pan-STARRS: 3π, yAB~20.5-21.0
• Dark Energy Survey, HSC, Euclid, LSST, …
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
12
Quasar searches in near IR surveys 68 z>5.5 quasars SDSS / CFHQS / Stripe82
Fan+ 2001-2006 Willott+ 2006-2010 Jiang+ 2008-09,2015
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
13
Quasar searches in near IR surveys 164 z>5.5 quasars
Mortlock+ 2011 Morganson+ 2012 Venemans+ 2013, 2015ab Bañados+ 2014, 2015 PhD thesis E. Bañados
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
UKIDSS / VIKING+KiDS / PS1
14
Quasar searches in near IR surveys 164 z>5.5 quasars SDSS / CFHQS / Stripe82 UKIDSS / VIKING / KiDS / PS1
Mortlock+ 2011 Venemans+ 2013, 2015a Bañados+ 2015
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
15
Quasar searches in near IR surveys
12 quasars confirmed at z≳6.5 1 quasar per >200 deg2
Redshift
Abs
olut
e U
V m
agni
tude
DES+VHS
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
←i-dropout limit
16
UKIDSS: a luminous quasar at z = 7.1
Bright source: M1450 = -26.6 (KAB = 19.6!)
1 2
Flux den
sity
Wavelength (micron)
Mortlock+ 2011
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
17
UKIDSS: a luminous quasar at z = 7.1
MgII line width + quasar luminosity: black hole mass
1 2
Flux den
sity
Wavelength (micron)
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
18
UKIDSS: a luminous quasar at z = 7.1
Emission lines: enrichment of broad line region
1 2
Flux den
sity
Wavelength (micron)
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
19
UKIDSS: a luminous quasar at z = 7.1
Absorption lines: reionisation, metals in IGM
1 2
Flux den
sity
Wavelength (micron)
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
20
UKIDSS: a luminous quasar at z = 7.1
1 2
Flux den
sity
Wavelength (micron)
Host galaxy: massive galaxy in formation (need to go to (sub)mm wavelengths)
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
21
Flux
den
sity
Observed wavelength
Black hole mass:
1.5x109 M⊙
1.0x109 M⊙
2.1x109 M⊙
e.g., De Rosa+ 2014
z = 6.9
z = 6.8
z = 6.6
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
22
RedshiO
Lbol = 0.5 x LEdd
1010
108
106
104 Lbol = LEdd
Bla
ck h
ole
mas
s
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
23
Enrichment of broad line region The Astrophysical Journal, 790:145 (14pp), 2014 August 1 De Rosa et al.
Figure 11. Fe ii/Mg ii line ratio as a function of redshift for z > 4. Black filledcircles: z > 6.5 sample; for J0109–3047 we are reporting the 3σ confidencelevel. Dark gray empty squares: 4.0 < z < 6.4 sample, De Rosa et al. (2011).Light gray empty squares: Fe ii/Mg ii ratios obtained for the z > 6.5 sampleusing the same continuum model as De Rosa et al. (2011). There is no evidencefor evolution of the estimated Fe ii/Mg ii line ratio as a function of cosmic agefor 4.0 < z < 7.1 in the quasar sample.
Fe ii/Mg ii ratio is not only sensitive to the corresponding Feand Mg abundances, but also to the hydrogen density, to theproperties of the radiation field and to the gas micro-turbulence(Baldwin et al. 2004; Verner et al. 2003, 2004; Bruhweiler &Verner 2008). Therefore, following De Rosa et al. (2011), wecan conclude that the lack of evolution in the Fe ii/Mg ii can beinterpreted as an early enrichment of the quasar host only underthe assumption that, in the analyzed sources, the physical con-ditions that determine the Fe ii emission are sufficiently similar.Under such assumption, the quasar hosts must have undergone amajor episode of Fe enrichment before the cosmic age at whichthey have been observed (∼0.8 Gyr). On the other hand, if Feand Mg are produced respectively via SNe Ia and core collapsesupernovae, one would expect Fe to be substantially produced atleast 1 Gyr after the initial burst of star formation (e.g., Hamann& Ferland 1993). However, the expected enrichment time andour observations are not in disagreement if we consider thatthe actual picture is probably more complex. Fe could in factbe generated by Pop III stars: extremely metal poor stars withtypical masses M ! 100 M⊙ that might be able to produce largeamounts of Fe by z " 10 (Heger & Woosley 2002). At the sametime, stellar nuclear yields are still rather uncertain, and thereare various scenarios in which significant metal production canoccur at early enough times to obtain a fully enriched BLR atz ∼ 7 (Matteucci & Recchi 2001; Venkatesan et al. 2004).
5. SUMMARY
We have analyzed of optical and NIR spectra of the onlyfour z > 6.5 quasars known to date: J1120+0641, discovered inthe UKIDSS-LAS survey, and J2348–3054, J0109–3047, andJ0305–3150, recently discovered in the VISTA-VIKING sur-vey. We presented new deep VLT/X-Shooter observations forJ1120+0641. Together with the new data, we analyzed all the
observations of the z > 6.5 sources collected by our group usingthe VLT/X-Shooter spectrograph and the Magellan/FIRE spec-trograph. The collected spectra provide essentially simultaneouscoverage of the 10000–24000 Å wavelength range. The qualityof this data set is likely the best achievable with the currentlyavailable facilities.
We used the spectra to estimate the masses of the BHs thatare powering these z > 6.5 quasars and to study their emission-line properties. The spectra were modeled using a combinationof a power-law continuum, a Balmer continuum, an Fe ii+Fe iii template, and a series of emission lines. We developeda maximum likelihood procedure for the spectral modeling,which allows a reliable estimate of how the uncertainties in thecontinuum modeling propagate into the estimates of the physicalquantities of interest. The z > 6.5 quasars are observationallyindistinguishable from their counterparts at lower redshifts.
We estimated the MBH from the Mg ii and C iv emission linesusing empirical mass-scaling relations. The MBH obtained fromthe two estimators agree within 1σ . The quasars in our samplehost BHs with masses of ∼109 M⊙ that are accreting closeto the Eddington luminosity (⟨log(LBol/LEdd)⟩ = −0.4 ± 0.2),in agreement with the average Eddington ratio obtained for a4.0 < z < 6.5 sample. If the measured average Eddington ratiois representative of the typical quasar accretion rate in the earlyuniverse, highly massive BH seeds (M0 ! 104 M⊙) need to bein place at very early times (z ! 20–30) in order to be able toobserve BHs with MBH ∼ 109 M⊙ at z = 6.5–7.1. At the sametime, quasar accretion episodes characterized by such high ratesmust be short in time and limited in number. If the BHs poweringour sources had been accreting at an Eddington ratio of ∼0.4for one additional characteristic accretion time (t ∼ 0.45 Gyr),they would have reached masses of ∼1011 M⊙, which is oneorder of magnitude larger than the most massive BH observedin the local universe.
We estimated the Si iv/C iv and C iii]/C iv flux ratios andcompared them with the results obtained from luminositymatched sub-samples at z ∼ 6 and 2 " z " 4.5. We findno evidence of evolution of these line ratios with cosmic time.
We calculated fluxes for the Mg ii and Fe ii lines and comparedthe measured Fe ii/Mg ii ratio with the results obtained for asample of 4.0 < z < 6.4 quasars. Since the Fe ii/Mg ii lineratio measurements are significantly dependent on the adoptedmodeling procedure, we performed a consistent analysis ofthe two samples. We do not detect any redshift evolution ofthe Fe ii/Mg ii ratio for 4.0 < z < 7.1. If we assume thatthe Fe ii/Mg ii line ratio is a reliable proxy of the Fe/Mgabundance ratio, this indicates that the z > 6.5 quasar hostsmust have undergone a major episode of Fe enrichment in thefirst ∼0.8 Gyr after the Big Bang.
We thank the referee for useful comments that allowed usto improve the quality of the paper. G.D.R. and B.M.P. aregrateful to the National Science Foundation for support of thiswork through grant AST-1008882 to The Ohio State University.B.P.V. acknowledges funding through the ERC grant “CosmicDawn”
REFERENCES
Arnaboldi, M., Neeser, M. J., Parker, L. C., et al. 2007, Msngr, 127, 28Assef, R. J., Frank, S., Grier, C. J., et al. 2012, ApJ, 753, 2Baldwin, J. A. 1977, ApJ, 214, 679Baldwin, J. A., Ferland, G. J., Korista, K. T., Hamann, F., & LaCluyze, A.
2004, ApJ, 615, 610
13
FeII / M
gII
z
No evolution of the enrichment close to the BH up to z ~ 7
e.g., De Rosa+ 2014
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
24
Quasar ionisation region
Near zone size: R ~ fHI-1/3 (e.g. Fan+ 2006)
1 2
Flux den
sity
Wavelength (micron)
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
25
Quasar ionisation regions
Redshift
Qua
sar n
ear z
one
[Mpc
] Near zone sizes scaled to same luminosity
Venemans+ 2015a
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
26
Quasar ionisation regions
Redshift
Qua
sar n
ear z
one
[Mpc
] ⇒ fHI increases by a factor of ~180 at 6<z<7 ⇒ fHI > few % at z=7 More work needed!
But see also: Bolton & Haehnelt 2007 Keating+ 2015
Near zone factor 6.5↓
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
27
Characteristics of distant quasars
• Due to bright central source, need to go to far-infrared to detect host galaxy
• PdBI DDT of two quasars at z = 6.5 and z = 7.1
• ALMA Cycle 1 observations of 3 z>6.6 quasars
→ Snapshot: ~15 minutes on-source
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
28
ALMA imaging of 6.6<z<6.9 quasars 15 min on-source (Cycle 1) [CII] emission line & FIR dust continuum detected: - L[CII] = 2-4 x109 L⊙ - LFIR = 1-8x1012 L⊙ → SFR~100-1500 M⊙/yr
Line widths: 250 – 400 km/s
6 6
4 2 0
Flux density [mJy]
0
0
6
20
z = 6.90
z = 6.79
z = 6.61
10
29
Map of [CII] emission of z=6.6 host
Continuum-subtracted map: 25sigma detection
Emission marginally resolved in 0.6˝ beam Size: 3.3x2.3 kpc
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
30
2hr PdBI on MUV=-27.2 quasar at z=6.5
Brightest [CII] emitter at z>6.4 Tentative detection of H2O Ideal to study ISM with ALMA
[CII]
H20(?)
Bañados+ 2015b
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching
31
Summary & outlook
• Quasars ideal targets to study the early universe
• Need large surveys to discover them • Supermassive (>109 Msun) BHs in 2-3 kpc dusty
galaxies forming stars at >100 Msun/yr
• Constrain neutral HI fraction and metal enrichment in IGM
• Study the ISM in massive galaxies up to z>7 (ALMA C3 project to study 35 quasars at z>5.9)
06/10/2015 The most distant quasars, Rainbows 2015, ESO Garching